By Nicolas Genin from Paris, France (66ème Festival de Venise (Mostra)) [CC BY-SA 2.0 (http://creativecommons.org/licenses/by-sa/2.0)], via Wikimedia Commons
July 27, 2017 Comments Closed

Avoiding the 7 Deadly Sins

(and the swimming pool)

We love Nicolas Cage. His resume goes back to the early ‘80s and includes more than 75 motion pictures, including an Academy Award for Leaving Las Vegas in 1985.

Did you know there is an ominous correlation between the number of Americans who drown each year and the number of films Cage appears in? It’s true. See here: http://tylervigen.com/view_correlation?id=359.

Obviously, these two facts are completely unrelated. Nick has been in some bad movies, but not bad enough for people to drown themselves.

The relationship between Cage films and drowning is what’s called a “spurious correlation:” just because two things graph together doesn’t mean there is a causal relationship between them.

In the big data, computer-driven world we live in, such correlations are easy to make (especially for the novice). The book Freakonomics, from 2005, turned finding these odd relationships into a cottage industry, of sorts.

Unfortunately, correlations aren’t always causal. And misinterpreting them as such can lead researchers, marketers and brand managers astray.

Spurious correlations are just one example of data interpretation mistakes that are the focal point of an interesting and amusing article titled The Seven Deadly Sins of Statistical Misinterpretation, and How to Avoid Them by Winnifred Louis and Cassandra Chapman. They are researchers at The University of Queensland in Australia.

In their article, Louis and Chapman look at common mistakes data researchers make and how to avoid them when it comes to statistics, probability and risk.

Their top 7 are:

  • Assuming small differences are meaningful
  • Equating statistical significance with real-world significance
  • Neglecting to look at extremes
  • Trusting coincidence (this is where the Cage example fits in)
  • Getting causation backwards
  • Forgetting to consider outside causes
  • Deceptive graphs

Here at Persuadable Research, we take data very seriously and we found this article insightful, as well as cautionary. Cautionary, because the data interpretation pitfalls Louis and Chapman point out are very real — and potentially very expensive.

Bad interpretation can lead to costly mistakes.

Our experience has taught us to be on the lookout for these data interpretation traps. It’s one of the advantages our clients get when they hire us:  a seasoned crew of researchers and analysts who know when something matters… and when it doesn’t.

See you by the pool!

 

Nicolas Cage image by Nicolas Genin from Paris, France (66ème Festival de Venise (Mostra)) [CC BY-SA 2.0 (http://creativecommons.org/licenses/by-sa/2.0)], via Wikimedia Commons

Copyright © 2017 Persuadable Research Corp.